Newsgroups: sci.lang
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!usenet.eel.ufl.edu!news-feed-1.peachnet.edu!gatech!swrinde!pipex!uknet!festival!edcogsci!iad
From: iad@cogsci.ed.ac.uk (Ivan A Derzhanski)
Subject: Re: Is Turkish a new language?
Message-ID: <D4nnFr.2tL@cogsci.ed.ac.uk>
Organization: Centre for Cognitive Science, Edinburgh, UK
References: <D4Hr7t.MC9@midway.uchicago.edu> <3il14u$ll7@sarasvati.umiacs.umd.edu> <D4MDKL.K96@midway.uchicago.edu>
Date: Mon, 27 Feb 1995 10:50:13 GMT
Lines: 71

In article <D4MDKL.K96@midway.uchicago.edu> deb5@midway.uchicago.edu writes:
>In article <3il14u$ll7@sarasvati.umiacs.umd.edu>, Suleyman Cenk Sahinalp <jenk@umiacs.umd.edu> wrote:
>>A very popular example demostrating this fact is:
>>   "Evcillestiremediklerimizden misiniz?" 
>>which is a one word sentence meaning:
>>   "Are you one of many, whom we couldn't domesticate?"
[...]
>Of course, this is an extreme example, like Shaw's "ghoti" or German 
>Schuetzengrabenvernichtungsmaschine.

It is most emphatically nothing like Shaw's *_ghoti_.  *_Ghoti_ is not
a legitimate spelling of an English word, whereas _evcille$tiremedik-
lerimizden misiniz_ is a legitimate Turkish word.

(And yes, it is a single word, despite the space before _mi-_, because
the front vowel in _ev_ determines the frontness of the vowels all the
way up to the ones in _-siniz_.)

It is an extreme example in that words of that length are rather rare,
but the only reason for that seems to be that it combines a long stem
(_evcille$tir-_) with a long sequence of inflexional affixes.

>The real advantage to Turkish is its regularity:  all the suffixes go
>on in a fixed order and have, at most, four completely predictable
>forms each.

... at most sixteen forms each.  Take the nominaliser _-dik_ from the
example above.  Its form is actually _-d/t-i/i"/u/u"-k/g^-_, depending
on what comes before, what (if anything) comes after and vowel harmony.
2 * 4 * 2 = 16.  But you're right that they are completely predictable.

(It is ironic that, when the agglutinative character of Hungarian is
discussed, people like talking _elo"rela'thatatlansa'gairo'l_ `of its
unforeseeabilities'.  :-)

  _elo"re-_ `fore-'
  _la't_    `see'
  _-hat_    potential
  _-atlan_  privative
  _-sa'g_   nominalisation
  _-a_      3Sg possessor
  _-i_      plural possessum
  _-ro'l_   delative

Anyone have a Basque word of that sort ready?  :-))

>Comparing one language to others, though, is difficult, because the
>single most important factor for determining how difficult a language
>is to learn is how it compares to your own.

I disagree.

>It was easier for me to learn German than Korean because I'm a native
>speaker of English; for a Japanese, the opposite would be true.

I found Hungarian much easier than German or French, although I'm not
a native speaker of an Uralic language.

>As far as Semitic goes, its paradigms appear rather regular, from
>what I'm told.  I've even heard Arabs brag that theirs is the best
>language for logicians because it's so regularly structured.

It does leave a few things to be desired.  For example, the plural
form(s) of a noun and the ma.sdar of a verb are unpredictable in most
cases, and have to be memorised.

-- 
`I'm sendin a flood tae pit an end tae it aw.  But dinny worry yersel, Noah.'
Ivan A Derzhanski (iad@cogsci.ed.ac.uk)    (J Stuart, _Auld Testament Tales_)
* Centre for Cognitive Science,  2 Buccleuch Place,   Edinburgh EH8 9LW,  UK
* Cowan House E113, Pollock Halls, 18 Holyrood Pk Rd, Edinburgh EH16 5BD, UK
